default_statistics_target参数对PG和MogDB性能影响测试和分析
default_statistics_target 参数对 PG 和 MogDB 性能影响测试和分析
本文出处：https://www.modb.pro/db/230160
前段时间在某客户生产环境优化 PG 查询性能时发现一个很有意思的现象，一个统计信息相关参数可以很大程度上影响 SQL 的性能。近期在学习 MogDB，就对比 PG 和 MogDB 做了个小测试 。
上图是对有性能问题的 SQL 做分析的截图，可以看到一个很奇怪的现象 Planning Time 占比较大，甚至远远大于具体执行的时间 Execution Time。
参考官方文档对 Planning Time 的解释’The Planning time shown by EXPLAIN ANALYZE is the time it took to generate the query plan from the parsed query and optimize it. It does not include parsing or rewriting.’。这说明生成查询计划和优化花费了较多的时间。
经过排查我们发现 Planning Time 耗时较长是由于系统参数 default_statistics_target 被设置成 10000 导致的。default_statistics_target 是用来控制默认统计信息目标的参数，默认值 100、最大值 10000，也可以建表时使用SET STATISTICS为每一列指定不同的值。 参考官方文档的描述Larger values increase the time needed to do ANALYZE, but might improve the quality of the planner's estimates , 增加这个参数会这增加 analyze 的时间，但能提升执行计划的质量。但在现实的场景中事实上上对 analyze 和 查询性能都会造成较大的影响。
简单分析下原因。
参数 default_statistics_target 会直接影响 Analyze 时采样的默认样本数。在 Analyze 时默认取 300*default_statistics_target 个随机块，再使用 Vitter 算法创建随机样本。
postgres=# select name,setting from pg_settings where name ='default_statistics_target';
name | setting
---------------------------+---------
default_statistics_target | 100
(1 row)
[omm@mogdb ~]$ gs_guc set -I all -N all -c "default_statistics_target=10000"
Begin to perform the total nodes: 3.
Popen count is 3, Popen success count is 3, Popen failure count is 0.
Begin to perform gs_guc for datanodes.
Command count is 3, Command success count is 3, Command failure count is 0.
Total instances: 3. Failed instances: 0.
ALL: Success to perform gs_guc!
[omm@mogdb ~]$ gs_om -t restart
postgres=# select name,setting from pg_settings where name ='default_statistics_target';
name | setting
---------------------------+---------
default_statistics_target | 10000
(1 row)
并且 default_statistics_target 也会影响统计信息中直方图的桶数。如果 numhist 的值大于 default_statistics_target，则按 default_statistics_target 统计。 直方图的桶数越多，就会影响优化器的执行效率。（具体原因待深入分析）
num_mcv = stats->attr->attstattarget;
num_bins = stats->attr->attstattarget;
num_hist = ndistinct - num_mcv;
if (num_hist > num_bins)
num_hist = num_bins + 1;
if (num_hist > num_bins)
num_hist = num_bins + 1;
另外为什么要乘以 300，注释中也给出了相关的解释 因为使用了 log 函数，所以表的大小对结果的影响较少，300*直方图的数量能覆盖大多数的需求。
/-------------------- * The following choice of minrows is based on the paper * "Random sampling for histogram construction: how much is enough?" * by Surajit Chaudhuri, Rajeev Motwani and Vivek Narasayya, in * Proceedings of ACM SIGMOD International Conference on Management * of Data, 1998, Pages 436-447. Their Corollary 1 to Theorem 5 * says that for table size n, histogram size k, maximum relative * error in bin size f, and error probability gamma, the minimum * random sample size is * r = 4 * k * ln(2n/gamma) / f^2 * Taking f = 0.5, gamma = 0.01, n = 10^6 rows, we obtain * r = 305.82 * k * Note that because of the log function, the dependence on n is * quite weak; even at n = 10^12, a 300*k sample gives <= 0.66 * bin size error with probability 0.99. So there's no real need to * scale for n, which is a good thing because we don't necessarily * know it at this point. *-------------------- */
在 PG 测试环境设置 default_statistics_target 参数分别为默认的 100 和最大值 10000，使用 BenchmarkSQL 在相同的测试参数下做 TPCC 测试，观察该参数对 tpmC 的影响。
Postgresql
PG 版本为 12.7
tpcc_db=# select version(); version --------------------------------------------------------------------------------------------------------- PostgreSQL 12.7 on x86_64-pc-linux-gnu, compiled by gcc (GCC) 4.8.5 20150623 (Red Hat 4.8.5-44), 64-bit (1 row)
加载数据并测试
[root@mogdb run]# sh runDatabaseBuild.sh props.postgres
[root@mogdb run]# sh runBenchmark.sh props.postgres
16:11:48,303 [main] INFO jTPCC : Term-00,
16:11:48,305 [main] INFO jTPCC : Term-00, +-------------------------------------------------------------+
16:11:48,305 [main] INFO jTPCC : Term-00, BenchmarkSQL v5.0
16:11:48,305 [main] INFO jTPCC : Term-00, +-------------------------------------------------------------+
16:11:48,305 [main] INFO jTPCC : Term-00, (c) 2003, Raul Barbosa
16:11:48,305 [main] INFO jTPCC : Term-00, (c) 2004-2016, Denis Lussier
16:11:48,307 [main] INFO jTPCC : Term-00, (c) 2016, Jan Wieck
16:11:48,307 [main] INFO jTPCC : Term-00, +-------------------------------------------------------------+
16:11:48,307 [main] INFO jTPCC : Term-00,
16:11:48,307 [main] INFO jTPCC : Term-00, db=postgres
16:11:48,307 [main] INFO jTPCC : Term-00, driver=org.postgresql.Driver
16:11:48,307 [main] INFO jTPCC : Term-00, conn=jdbc:postgresql://...:5432/tpcc_db?prepareThreshold=1&batchMode=on&fetchsize=10&loggerLevel=off&autoReconnect=true
16:11:48,308 [main] INFO jTPCC : Term-00, user=tpcc
16:11:48,308 [main] INFO jTPCC : Term-00,
16:11:48,308 [main] INFO jTPCC : Term-00, warehouses=10
16:11:48,308 [main] INFO jTPCC : Term-00, terminals=20
16:11:48,309 [main] INFO jTPCC : Term-00, runMins=3
16:11:48,309 [main] INFO jTPCC : Term-00, limitTxnsPerMin=0
16:11:48,309 [main] INFO jTPCC : Term-00, terminalWarehouseFixed=false
16:11:48,309 [main] INFO jTPCC : Term-00,
16:11:48,309 [main] INFO jTPCC : Term-00, newOrderWeight=10
16:11:48,309 [main] INFO jTPCC : Term-00, paymentWeight=10
16:11:48,309 [main] INFO jTPCC : Term-00, orderStatusWeight=60
16:11:48,309 [main] INFO jTPCC : Term-00, deliveryWeight=10
16:11:48,309 [main] INFO jTPCC : Term-00, stockLevelWeight=10
16:11:48,310 [main] INFO jTPCC : Term-00,
16:11:48,310 [main] INFO jTPCC : Term-00, resultDirectory=mogdb_result_%tY-%tm-%td_%tH%tM%tS
16:11:48,310 [main] INFO jTPCC : Term-00, osCollectorScript=null
16:11:48,310 [main] INFO jTPCC : Term-00,
16:11:48,321 [main] INFO jTPCC : Term-00, copied props.postgres to mogdb_result_2021-12-31_161148/run.properties
16:11:48,322 [main] INFO jTPCC : Term-00, created mogdb_result_2021-12-31_161148/data/runInfo.csv for runID 15
16:11:48,322 [main] INFO jTPCC : Term-00, writing per transaction results to mogdb_result_2021-12-31_161148/data/result.csv
16:11:48,322 [main] INFO jTPCC : Term-00,
16:11:48,404 [main] INFO jTPCC : Term-00, C value for C_LAST during load: 204
16:11:48,404 [main] INFO jTPCC : Term-00, C value for C_LAST this run: 126
16:11:48,405 [main] INFO jTPCC : Term-00,
Term-00, Running Average tpmTOTAL: 92146.82 Current tpmTOTAL: 2988612 Memory Usage: 59MB / 105MB
16:14:48,587 [Thread-5] INFO jTPCC : Term-00,
16:14:48,588 [Thread-5] INFO jTPCC : Term-00,
16:14:48,588 [Thread-5] INFO jTPCC : Term-00, Measured tpmC (NewOrders) = 9195.03
16:14:48,588 [Thread-5] INFO jTPCC : Term-00, Measured tpmTOTAL = 92130.3
16:14:48,588 [Thread-5] INFO jTPCC : Term-00, Session Start = 2021-12-31 16:11:48
16:14:48,588 [Thread-5] INFO jTPCC : Term-00, Session End = 2021-12-31 16:14:48
16:14:48,588 [Thread-5] INFO jTPCC : Term-00, Transaction Count = 276459
16:14:48,588 [Thread-5] INFO jTPCC : executeTime[Payment]=339035
16:14:48,588 [Thread-5] INFO jTPCC : executeTime[Order-Status]=1111348
16:14:48,588 [Thread-5] INFO jTPCC : executeTime[Delivery]=883458
16:14:48,589 [Thread-5] INFO jTPCC : executeTime[Stock-Level]=588809
16:14:48,589 [Thread-5] INFO jTPCC : executeTime[New-Order]=676662
清理测试数据
[root@mogdb run]# sh runDatabaseDestroy.sh props.postgres
------------------------------------------------------------
Loading SQL file ./sql.common/tableDrops.sql
------------------------------------------------------------
drop table bmsql_config;
drop table bmsql_new_order;
drop table bmsql_order_line;
drop table bmsql_oorder;
drop table bmsql_history;
drop table bmsql_customer;
drop table bmsql_stock;
drop table bmsql_item;
drop table bmsql_district;
drop table bmsql_warehouse;
drop sequence bmsql_hist_id_seq;
修改 PG 的配置，将 default_statistics_target 从默认的 100 改成 10000；
tpcc_db=# select name,setting from pg_settings where name ='default_statistics_target';
name | setting
---------------------------+---------
default_statistics_target | 100
(1 row)
tpcc_db=# alter system set default_statistics_target=10000;
ALTER SYSTEM
tpcc_db=# select pg_reload_conf();
pg_reload_conf
t
(1 row)
tpcc_db=# select name,setting from pg_settings where name ='default_statistics_target';
name | setting
---------------------------+---------
default_statistics_target | 10000
(1 row)
再次测试
[root@mogdb run]# sh runDatabaseBuild.sh props.postgres
[root@mogdb run]# sh runBenchmark.sh props.postgres
16:03:00,023 [main] INFO jTPCC : Term-00,
16:03:00,025 [main] INFO jTPCC : Term-00, +-------------------------------------------------------------+
16:03:00,026 [main] INFO jTPCC : Term-00, BenchmarkSQL v5.0
16:03:00,026 [main] INFO jTPCC : Term-00, +-------------------------------------------------------------+
16:03:00,026 [main] INFO jTPCC : Term-00, (c) 2003, Raul Barbosa
16:03:00,026 [main] INFO jTPCC : Term-00, (c) 2004-2016, Denis Lussier
16:03:00,028 [main] INFO jTPCC : Term-00, (c) 2016, Jan Wieck
16:03:00,028 [main] INFO jTPCC : Term-00, +-------------------------------------------------------------+
16:03:00,028 [main] INFO jTPCC : Term-00,
16:03:00,029 [main] INFO jTPCC : Term-00, db=postgres
16:03:00,029 [main] INFO jTPCC : Term-00, driver=org.postgresql.Driver
16:03:00,029 [main] INFO jTPCC : Term-00, conn=jdbc:postgresql://...:5432/tpcc_db?prepareThreshold=1&batchMode=on&fetchsize=10&loggerLevel=off&autoReconnect=true
16:03:00,029 [main] INFO jTPCC : Term-00, user=tpcc
16:03:00,029 [main] INFO jTPCC : Term-00,
16:03:00,029 [main] INFO jTPCC : Term-00, warehouses=10
16:03:00,029 [main] INFO jTPCC : Term-00, terminals=20
16:03:00,031 [main] INFO jTPCC : Term-00, runMins=3
16:03:00,031 [main] INFO jTPCC : Term-00, limitTxnsPerMin=0
16:03:00,031 [main] INFO jTPCC : Term-00, terminalWarehouseFixed=false
16:03:00,031 [main] INFO jTPCC : Term-00,
16:03:00,031 [main] INFO jTPCC : Term-00, newOrderWeight=10
16:03:00,031 [main] INFO jTPCC : Term-00, paymentWeight=10
16:03:00,031 [main] INFO jTPCC : Term-00, orderStatusWeight=60
16:03:00,031 [main] INFO jTPCC : Term-00, deliveryWeight=10
16:03:00,032 [main] INFO jTPCC : Term-00, stockLevelWeight=10
16:03:00,032 [main] INFO jTPCC : Term-00,
16:03:00,032 [main] INFO jTPCC : Term-00, resultDirectory=mogdb_result_%tY-%tm-%td_%tH%tM%tS
16:03:00,032 [main] INFO jTPCC : Term-00, osCollectorScript=null
16:03:00,032 [main] INFO jTPCC : Term-00,
16:03:00,045 [main] INFO jTPCC : Term-00, copied props.postgres to mogdb_result_2021-12-31_160300/run.properties
16:03:00,045 [main] INFO jTPCC : Term-00, created mogdb_result_2021-12-31_160300/data/runInfo.csv for runID 14
16:03:00,045 [main] INFO jTPCC : Term-00, writing per transaction results to mogdb_result_2021-12-31_160300/data/result.csv
16:03:00,045 [main] INFO jTPCC : Term-00,
16:03:00,132 [main] INFO jTPCC : Term-00, C value for C_LAST during load: 109
16:03:00,132 [main] INFO jTPCC : Term-00, C value for C_LAST this run: 211
16:03:00,132 [main] INFO jTPCC : Term-00,
Term-00, Running Average tpmTOTAL: 20052.14 Current tpmTOTAL: 664776 Memory Usage: 21MB / 105MB
16:06:04,432 [Thread-10] INFO jTPCC : Term-00,
16:06:04,432 [Thread-10] INFO jTPCC : Term-00,
16:06:04,433 [Thread-10] INFO jTPCC : Term-00, Measured tpmC (NewOrders) = 1997.57
16:06:04,433 [Thread-10] INFO jTPCC : Term-00, Measured tpmTOTAL = 20044.52
16:06:04,433 [Thread-10] INFO jTPCC : Term-00, Session Start = 2021-12-31 16:03:00
16:06:04,433 [Thread-10] INFO jTPCC : Term-00, Session End = 2021-12-31 16:06:04
16:06:04,433 [Thread-10] INFO jTPCC : Term-00, Transaction Count = 61530
16:06:04,434 [Thread-10] INFO jTPCC : executeTime[Payment]=79515
16:06:04,434 [Thread-10] INFO jTPCC : executeTime[Order-Status]=241522
16:06:04,434 [Thread-10] INFO jTPCC : executeTime[Delivery]=299391
16:06:04,434 [Thread-10] INFO jTPCC : executeTime[Stock-Level]=2863537
16:06:04,434 [Thread-10] INFO jTPCC : executeTime[New-Order]=165739
从测试结果看 default_statistics_target=10000 配置下 tpmC 只有默认配置下的 22.2%，TPS 和事务的延迟也有明显差距。实验证明增大该参数会对实例的性能造成较大影响。
综上，从原理和实验看在 PG 中 default_statistics_target 的值不要设置的太大，默认值 100 应该还是一个不错的选择。
MogDB
近期在测试 MogDB，很好奇修改这个参数是否对 MogDB 的性能也有很大的影响，在测试环境简单跑了下 TPCC 测试。【本地测试环境硬件性能比 PG 差很多，tpmC 结果不要横向比较】同样设置 default_statistics_target 参数分别为默认的 100 和最大值 10000。
[root@mogdb run]# cd /root/benchmarksql/run/
[root@mogdb run]# sh runDatabaseBuild.sh props.mogdb
------------------------------------------------------------
Loading SQL file ./sql.postgres/buildFinish.sql
------------------------------------------------------------
-- Extra commands to run after the tables are created, loaded,
-- indexes built and extra's created.
-- PostgreSQL version.
vacuum analyze;
启动测试
[root@mogdb run]# sh runBenchmark.sh props.mogdb
16:32:08,768 [main] INFO jTPCC : Term-00,
16:32:08,772 [main] INFO jTPCC : Term-00, +-------------------------------------------------------------+
16:32:08,772 [main] INFO jTPCC : Term-00, BenchmarkSQL v5.0
16:32:08,772 [main] INFO jTPCC : Term-00, +-------------------------------------------------------------+
16:32:08,772 [main] INFO jTPCC : Term-00, (c) 2003, Raul Barbosa
16:32:08,772 [main] INFO jTPCC : Term-00, (c) 2004-2016, Denis Lussier
16:32:08,776 [main] INFO jTPCC : Term-00, (c) 2016, Jan Wieck
16:32:08,777 [main] INFO jTPCC : Term-00, +-------------------------------------------------------------+
16:32:08,777 [main] INFO jTPCC : Term-00,
16:32:08,777 [main] INFO jTPCC : Term-00, db=postgres
16:32:08,777 [main] INFO jTPCC : Term-00, driver=org.postgresql.Driver
16:32:08,777 [main] INFO jTPCC : Term-00, conn=jdbc:postgresql://...:26000/tpcc_db?prepareThreshold=1&batchMode=on&fetchsize=10&loggerLevel=off&autoReconnect=true
16:32:08,777 [main] INFO jTPCC : Term-00, user=tpcc
16:32:08,777 [main] INFO jTPCC : Term-00,
16:32:08,777 [main] INFO jTPCC : Term-00, warehouses=10
16:32:08,777 [main] INFO jTPCC : Term-00, terminals=20
16:32:08,778 [main] INFO jTPCC : Term-00, runMins=3
16:32:08,779 [main] INFO jTPCC : Term-00, limitTxnsPerMin=0
16:32:08,779 [main] INFO jTPCC : Term-00, terminalWarehouseFixed=false
16:32:08,779 [main] INFO jTPCC : Term-00,
16:32:08,779 [main] INFO jTPCC : Term-00, newOrderWeight=10
16:32:08,779 [main] INFO jTPCC : Term-00, paymentWeight=10
16:32:08,779 [main] INFO jTPCC : Term-00, orderStatusWeight=60
16:32:08,779 [main] INFO jTPCC : Term-00, deliveryWeight=10
16:32:08,779 [main] INFO jTPCC : Term-00, stockLevelWeight=10
16:32:08,779 [main] INFO jTPCC : Term-00,
16:32:08,779 [main] INFO jTPCC : Term-00, resultDirectory=mogdb_result_%tY-%tm-%td_%tH%tM%tS
16:32:08,779 [main] INFO jTPCC : Term-00, osCollectorScript=null
16:32:08,779 [main] INFO jTPCC : Term-00,
16:32:08,796 [main] INFO jTPCC : Term-00, copied props.mogdb to mogdb_result_2021-12-31_163208/run.properties
16:32:08,796 [main] INFO jTPCC : Term-00, created mogdb_result_2021-12-31_163208/data/runInfo.csv for runID 16
16:32:08,796 [main] INFO jTPCC : Term-00, writing per transaction results to mogdb_result_2021-12-31_163208/data/result.csv
16:32:08,796 [main] INFO jTPCC : Term-00,
16:32:08,868 [main] INFO jTPCC : Term-00, C value for C_LAST during load: 36
16:32:08,869 [main] INFO jTPCC : Term-00, C value for C_LAST this run: 130
16:32:08,869 [main] INFO jTPCC : Term-00,
Term-00, Running Average tpmTOTAL: 36569.79 Current tpmTOTAL: 1185828 Memory Usage: 23MB / 105MB
16:35:09,089 [Thread-12] INFO jTPCC : Term-00,
16:35:09,089 [Thread-12] INFO jTPCC : Term-00,
16:35:09,089 [Thread-12] INFO jTPCC : Term-00, Measured tpmC (NewOrders) = 3651.92
16:35:09,094 [Thread-12] INFO jTPCC : Term-00, Measured tpmTOTAL = 36558.86
16:35:09,094 [Thread-12] INFO jTPCC : Term-00, Session Start = 2021-12-31 16:32:09
16:35:09,094 [Thread-12] INFO jTPCC : Term-00, Session End = 2021-12-31 16:35:09
16:35:09,094 [Thread-12] INFO jTPCC : Term-00, Transaction Count = 109728
16:35:09,094 [Thread-12] INFO jTPCC : executeTime[Payment]=377510
16:35:09,094 [Thread-12] INFO jTPCC : executeTime[Order-Status]=975192
16:35:09,094 [Thread-12] INFO jTPCC : executeTime[Delivery]=1057319
16:35:09,094 [Thread-12] INFO jTPCC : executeTime[Stock-Level]=557707
16:35:09,095 [Thread-12] INFO jTPCC : executeTime[New-Order]=631551
清理测试数据
[root@mogdb run]# sh runDatabaseDestroy.sh props.mogdb
------------------------------------------------------------
Loading SQL file ./sql.common/tableDrops.sql
------------------------------------------------------------
drop table bmsql_config;
drop table bmsql_new_order;
drop table bmsql_order_line;
drop table bmsql_oorder;
drop table bmsql_history;
drop table bmsql_customer;
drop table bmsql_stock;
drop table bmsql_item;
drop table bmsql_district;
drop table bmsql_warehouse;
drop sequence bmsql_hist_id_seq;
我们修改 MogDB 的配置，将 default_statistics_target 从默认的 100 改成 10000；
postgres=# select name,setting from pg_settings where name ='default_statistics_target';
name | setting
---------------------------+---------
default_statistics_target | 100
(1 row)
[omm@mogdb ~]$ gs_guc set -I all -N all -c "default_statistics_target=10000"
Begin to perform the total nodes: 3.
Popen count is 3, Popen success count is 3, Popen failure count is 0.
Begin to perform gs_guc for datanodes.
Command count is 3, Command success count is 3, Command failure count is 0.
Total instances: 3. Failed instances: 0.
ALL: Success to perform gs_guc!
[omm@mogdb ~]$ gs_om -t restart
postgres=# select name,setting from pg_settings where name ='default_statistics_target';
name | setting
---------------------------+---------
default_statistics_target | 10000
(1 row)
再次进行 benchmarksql 测试
[root@mogdb run]# sh runDatabaseBuild.sh props.mogdb
[root@mogdb run]# sh runBenchmark.sh props.mogdb
16:47:36,924 [main] INFO jTPCC : Term-00,
16:47:36,928 [main] INFO jTPCC : Term-00, +-------------------------------------------------------------+
16:47:36,928 [main] INFO jTPCC : Term-00, BenchmarkSQL v5.0
16:47:36,929 [main] INFO jTPCC : Term-00, +-------------------------------------------------------------+
16:47:36,929 [main] INFO jTPCC : Term-00, (c) 2003, Raul Barbosa
16:47:36,929 [main] INFO jTPCC : Term-00, (c) 2004-2016, Denis Lussier
16:47:36,935 [main] INFO jTPCC : Term-00, (c) 2016, Jan Wieck
16:47:36,935 [main] INFO jTPCC : Term-00, +-------------------------------------------------------------+
16:47:36,935 [main] INFO jTPCC : Term-00,
16:47:36,935 [main] INFO jTPCC : Term-00, db=postgres
16:47:36,936 [main] INFO jTPCC : Term-00, driver=org.postgresql.Driver
16:47:36,936 [main] INFO jTPCC : Term-00, conn=jdbc:postgresql://...:26000/tpcc_db?prepareThreshold=1&batchMode=on&fetchsize=10&loggerLevel=off&autoReconnect=true
16:47:36,936 [main] INFO jTPCC : Term-00, user=tpcc
16:47:36,936 [main] INFO jTPCC : Term-00,
16:47:36,936 [main] INFO jTPCC : Term-00, warehouses=10
16:47:36,936 [main] INFO jTPCC : Term-00, terminals=20
16:47:36,937 [main] INFO jTPCC : Term-00, runMins=3
16:47:36,937 [main] INFO jTPCC : Term-00, limitTxnsPerMin=0
16:47:36,938 [main] INFO jTPCC : Term-00, terminalWarehouseFixed=false
16:47:36,938 [main] INFO jTPCC : Term-00,
16:47:36,938 [main] INFO jTPCC : Term-00, newOrderWeight=10
16:47:36,938 [main] INFO jTPCC : Term-00, paymentWeight=10
16:47:36,938 [main] INFO jTPCC : Term-00, orderStatusWeight=60
16:47:36,938 [main] INFO jTPCC : Term-00, deliveryWeight=10
16:47:36,938 [main] INFO jTPCC : Term-00, stockLevelWeight=10
16:47:36,938 [main] INFO jTPCC : Term-00,
16:47:36,938 [main] INFO jTPCC : Term-00, resultDirectory=mogdb_result_%tY-%tm-%td_%tH%tM%tS
16:47:36,938 [main] INFO jTPCC : Term-00, osCollectorScript=null
16:47:36,938 [main] INFO jTPCC : Term-00,
16:47:36,954 [main] INFO jTPCC : Term-00, copied props.mogdb to mogdb_result_2021-12-31_164736/run.properties
16:47:36,954 [main] INFO jTPCC : Term-00, created mogdb_result_2021-12-31_164736/data/runInfo.csv for runID 17
16:47:36,955 [main] INFO jTPCC : Term-00, writing per transaction results to mogdb_result_2021-12-31_164736/data/result.csv
16:47:36,955 [main] INFO jTPCC : Term-00,
16:47:37,037 [main] INFO jTPCC : Term-00, C value for C_LAST during load: 201
16:47:37,037 [main] INFO jTPCC : Term-00, C value for C_LAST this run: 117
16:47:37,038 [main] INFO jTPCC : Term-00,
Term-00, Running Average tpmTOTAL: 33621.25 Current tpmTOTAL: 1088436 Memory Usage: 27MB / 105MB
16:50:37,412 [Thread-17] INFO jTPCC : Term-00,
16:50:37,413 [Thread-17] INFO jTPCC : Term-00,
16:50:37,413 [Thread-17] INFO jTPCC : Term-00, Measured tpmC (NewOrders) = 3404.5
16:50:37,413 [Thread-17] INFO jTPCC : Term-00, Measured tpmTOTAL = 33593.8
16:50:37,413 [Thread-17] INFO jTPCC : Term-00, Session Start = 2021-12-31 16:47:37
16:50:37,413 [Thread-17] INFO jTPCC : Term-00, Session End = 2021-12-31 16:50:37
16:50:37,413 [Thread-17] INFO jTPCC : Term-00, Transaction Count = 100884
16:50:37,414 [Thread-17] INFO jTPCC : executeTime[Payment]=336881
16:50:37,414 [Thread-17] INFO jTPCC : executeTime[Order-Status]=818460
16:50:37,414 [Thread-17] INFO jTPCC : executeTime[Delivery]=923761
16:50:37,414 [Thread-17] INFO jTPCC : executeTime[Stock-Level]=960065
16:50:37,414 [Thread-17] INFO jTPCC : executeTime[New-Order]=561205
可见 MogDB 经过优化，增加 default_statistics_target 参数对性能的影响可以忽略。小结综上，default_statistics_target 参数对 PG 性能影响较大，建议保持默认值。而 MogDB 经过优化可以选择较高的 default_statistics_target 值，这样可以记录更精确的统计信息用以获得更准确的执行计划。
default_statistics_target参数对PG和MogDB性能影响测试和分析的更多相关文章
- 对MySQL性能影响较大的五类配置参数
以下主要是对MySQL 性能影响关系紧密的五大配置参数的介绍. 一. 连接 连接通常来自Web 服务器,下面列出了一些与连接有关的参数,以及该如何设置它们. (一). ...
- (转)MySQL配置文件mysql.ini参数详解、MySQL性能优化
本文转自:http://www.cr173.com/html/18331_1.html my.ini(Linux系统下是my.cnf),当mysql服务器启动时它会读取这个文件,设置相关的运行环境参数 ...
- MySQL配置文件mysql.ini参数详解、MySQL性能优化
my.ini(Linux系统下是my.cnf),当mysql服务器启动时它会读取这个文件,设置相关的运行环境参数. my.ini分为两块:Client Section和Server Section. ...
- C#异常性能影响
何谓异常 很多人在讨论异常的时候很模糊,仿佛所谓异常就是try{}catch{},异常就是Exception,非常的片面,所以导致异常影响性能,XXXX……等很多奇怪的言论,所以在此我意在对异常正名. ...
- 盘点 Oracle 11g 中新特性带来的10大性能影响
Oracle的任何一个新版本,总是会带来大量引人瞩目的新特性,但是往往在这些新特性引入之初,首先引起的是一些麻烦,因为对于新技术的不了解.因为对于旧环境的不适应,从Oracle产品到技术服务运维,总是 ...
- GSO/TSO/GRO等对VirtIO虚机的网络性能影响分析(by quqi99)
作者:张华 发表于:2016-04-05版权声明:可以任意转载,转载时请务必以超链接形式标明文章原始出处和作者信息及本版权声明 ( http://blog.csdn.net/quqi99 ) IP层 ...
- 考查SQLite 3索引对整数排序的性能影响
做个实验,想了解SQLite3索引对整数排序的性能影响. 用这个测试表,考查绿色那列: id name date 自增型主键 字符串型,随机生成 整数型 随机生成,范围0到54354354 1 bMz ...
- Salesforce select字段的多少对性能影响巨大
Salesforce select字段的多少对性能影响巨大,第1个是select 144个字段,第2个是select 5个字段, 性能相差了7倍 "select Id,IsDeleted,M ...
- mysql中in和exists二者的区别和性能影响
mysql查询语句in和exists二者的区别和性能影响 还记得一次面试中被人问到in 和 exists的区别,当然只是草草做答,现在来做下分析. mysql中的in语句是把外表和内表作hash 连接 ...
- numa对MySQL多实例性能影响
numa对MySQL多实例性能影响,通过对numa将MySQL绑定在不同的CPU节点上,并且采用绑定的内存分配策略,强制在本节点内分配内存.具体测试如下:1.关闭numa(numa= interle ...
随机推荐
- Linux操作系统加固建议
1.1.1 口令锁定策略 1.执行备份 (1).redhat执行备份: #cp -p /etc/pam.d/system-auth /etc/pam.d/system-auth_bak (2).SUS ...
- 俄罗斯套娃 (Matryoshka) 嵌入模型概述
在这篇博客中,我们将向你介绍俄罗斯套娃嵌入的概念,并解释为什么它们很有用.我们将讨论这些模型在理论上是如何训练的,以及你如何使用 Sentence Transformers 来训练它们. 除此之外,我 ...
- MSYS2使用记录——win10系统64位安装msys2最新版（msys2-x86_64-20190524.exe）
MSYS2使用记录--win10系统64位安装msys2最新版(msys2-x86_64-20190524.exe) 安装 测试系统: windows 10 首先需要msys2的安装包,可以去官网下载 ...
- 摆脱鼠标系列 - vscode - ctrl+up 光标上移动4行 ctrl+down 光标下移4行
为什么 摆脱鼠标系列 - vscode - ctrl+up 光标上移动4行 之前滚动屏幕总是用鼠标,现在改为 ctrl + 上箭头或下箭头 实现起来稍微有些麻烦 实现 需要安装 macros 插件 这 ...
- 基于泰凌微的TLSR8355芯片的2.4G无线私有协议PCBA设计调试总结
一 前记 经常做物联网的类的产品,TLSR8355凭借着它的射频距离远,功能强大等优点成为很多客户的首选.TLSR8355系列专用于2.4GHz射频系统芯片解决方案,如零售/物流.专用网络.Beaco ...
- python 读取串口数据常用函数及实例分析
前记: 人生苦短,我用python,python在做一些算法验证和接口验证方面,的确是非常的好用.读取串口经常用到,这里就做个总结,给自己和周围的人做个备忘吧. 函数解析: 初始化串口数据: impo ...
- 快速将json装DTO的GsonFormatPlus插件使用
参考:https://www.jianshu.com/p/8fb0e4274436 https://blog.csdn.net/qq_43039260/article/details/12676582 ...
- RTP网络包提取码流方法
1.工具wireshark,官网最新的版本没有保存按键,建议使用3.4 链接:https://pan.baidu.com/s/1VHSGgojZPNtzyzV6RMVjkw 提取码:q51a 2.wi ...
- Vue3.0里为什么要用 Proxy API 替代 defineProperty API ？
这里给大家分享我在网上总结出来的一些知识,希望对大家有所帮助 一.Object.defineProperty 定义:Object.defineProperty() 方法会直接在一个对象上定义一个新属性 ...
- B站框框老师+宋浩老师概率论视频课笔记，宋浩老师数理统计视频课笔记
只做理解类记录,哪个知识点忘了去看视频.前四章是概率,看的框框老师. 概率论 1.随机试验:可重复性.可预知性.不确定性 2.样本空间:随机试验E的所有可能结果,记为S或Ω 3.样本点:样本空间中的每 ...